ChatGPT (GPT-6)
Good for agent work and coding on Plus with GPT-6 Astra. For writing, Gemini AI Pro and Claude Pro score higher.
Going away: ChatGPT Library (Pro, Business, Enterprise accounts with Space), no date given yet. What OpenAI offers instead: Space. Source: help.openai.com, read 19 Oct 2026.
Going away: Custom GPTs in ChatGPT (all plans), no date given yet. What OpenAI offers instead: plugins. Source: help.openai.com, read 19 Oct 2026.
Should you use it?
Pick it if
People who want one app for multi-step agent work and coding and are fine paying $20 a month.
Best for: People who want one app for agent tasks and coding on the $20 Plus plan (AI apps Top 20)
Skip it if
You mainly write: Gemini AI Pro and Claude Pro both score 70 for writing, against 50 here.
You give up: Writing scores 50
The short version
- What this is
- OpenAI's assistant app. Since 7 Oct 2026 chat runs on GPT-6 Sol for paying users and GPT-6 Luna on Free and Go.
- Why it matters
- It is the assistant most people start with; this shows where its paid plan leads and where others do better.
- Plus ($20/month) includes GPT-6 Astra in Work and Codex, 2nd on both the Agent and WebDev boards s1 s2 s15.
- Writing is its weak spot: GPT-6 Sol ranks 63rd of 412 on Creative Writing s4.
- Chats can train OpenAI models unless you switch it off in Data controls s9.
Good at
- GPT-6 Astra, included in Plus, ranks 2nd of 142 on arena.ai WebDev and 2nd of 52 on the Agent board s1 s2 s6.
- OpenAI's GPT-6 models hold the top four places on the arena.ai Agent Alignment Index, which counts unauthorised actions and false reports of finished work s14.
- One app covers chat, long agent tasks and coding, and you can export your data s6 s10.
Our take
ChatGPT Plus is a sound pick for agent tasks and coding. OpenAI says Plus includes GPT-6 Astra in ChatGPT Work and Codex s15, and Astra is 2nd on both the Agent and WebDev boards s1 s2, so under our rule, the best model each plan lets you choose, Plus scores 70 for both. OpenAI's GPT-6 models also lead the Agent Alignment Index, so they are among the least likely to act without permission or claim unfinished work is done s14.
Chat is weaker. Plus chat offers only GPT-6 Sol, at Medium and High thinking s15, and Sol is 44th on Text Expert and 63rd on Creative Writing s3 s4, so for hard questions and writing the free plan, on GPT-6 Luna, scores higher because it costs nothing. GPT-6.1 Sol ranks higher, but Plus gets it only in Work and Codex s15.
In a Tom's Guide face-off on 24 Sep, before chat moved to GPT-6 Sol, Claude Opus 5.5 won 4 of 5 prompts; the ChatGPT-6 version was not named s11. The ranks are for the highest effort settings listed, which Plus chat does not include, so treat each score as that plan's ceiling. In Codex, the default model depends on your setup s12.
Scores by job
A score from 0 to 100 per job, against everything available today. Overall 63 across 4 jobs, the average of the job scores.
New to this? What the scores and labels mean
- 0 to 100
- Our score for one job, like writing or coding, compared with everything else out today. 100 is the best available today. For AI models we build it from public data, such as Arena's blind votes and Epoch AI's independent tests; apps and tools are scored on our own checklist. The same tool can score differently for different jobs, and where two likely ranges overlap, we call it a tie.
- Our verdict
- Use it: a top choice for most people. Worth trying: good, and worth a go. Depends: good for some people or one kind of job. Wait: too early to tell, or not out for everyone yet. Skip: something else does the job better.
- First look
- Just released. We say what it is and whether to bother, and give it a score once public results exist.
- No score yet
- None of the public sources we use covers it yet, or the only evidence is from a source whose numbers we can't use, so we describe it in words instead.
- Best buy
- Nothing cheaper scores as well for that job.
- Best free
- The best score for that job that costs nothing, through a free plan or by running on your own computer.
- Pay as you go
- No monthly plan: you pay for each use, usually through a developer account.
- Checked
- The day we last confirmed the score and the price against the sources. Prices change often, so check the date.
- AI model
- The engine underneath, like GPT or Claude. You use it through an app or, if you code, directly.
- Open source
- Free to download and run on your own computer, so nothing leaves your machine.
Our checklist score, going by a public leaderboard and the maker’s pages.
Our checklist
1 source so far
Sources and notes
Public leaderboardMaker’s pagesFree (GPT-6 Luna, the best GPT-6 model Free lists, s6, s8; limited Codex and Work, s6): 6/10, does the job 1 (GPT 6 Luna (Max) arena.ai Agent rank 25 of 52, +1.93%, s2), underlying quality 1, value 2 ($0), privacy 1 (training unless you opt out, s9), limits 1 (not published). Plus (GPT-6 Astra, which OpenAI says Plus includes in ChatGPT Work and Codex, s15): 7/10, does the job 2 (GPT 6 Astra (Max) Agent rank 2 of 52, +13.09%, s2), underlying quality 2, value 1 ($20/month, s7), privacy 1, limits 1 (not published). Score from Plus, the higher total. Sources s2 s6 s15.
Our checklist score, going by a public leaderboard and the maker’s pages.
Our checklist
1 source so far
Sources and notes
Public leaderboardMaker’s pagesFree (GPT-6 Luna in chat, s8; Free Codex lists only GPT-5.6 Terra, s15): 6/10, does the job 1 (arena.ai WebDev rank 34 of 142, gpt-6-luna-max, s1), underlying quality 1, value 2 ($0), privacy 1 s9, limits 1. Plus (GPT-6 Astra in ChatGPT Work, s15): 7/10, does the job 2 (WebDev rank 2 of 142, gpt-6-astra-max, s1), underlying quality 2, value 1 ($20/month, s7), privacy 1, limits 1. Score from Plus, the higher total. Plus chat alone, on GPT-6 Sol, would score 5.
Our checklist score, going by a public leaderboard and the maker’s pages.
Our checklist
1 source so far
Sources and notes
Public leaderboardMaker’s pagesFree (GPT-6 Luna, s8): 6/10, does the job 1 (arena.ai Text Expert rank 56 of 364, gpt-6-luna-max, s3), underlying quality 1, value 2 ($0), privacy 1 s9, limits 1. Plus (GPT-6 Sol, the only GPT-6 model in Plus chat, s15): 5/10, does the job 1 (rank 44 of 364, gpt-6-sol-max, s3), underlying quality 1, value 1 ($20/month, s7), privacy 1, limits 1. Score from Free, the higher total. GPT-6.1 Sol, rank 8 s3, is offered to Plus only in Work and Codex s6 s15, which we count for agents and coding, not for answering questions in chat. Top plan: Pro ($100/month) adds GPT-6 Astra in chat, which would score 4 (value 0 at that price, s13, s15).
Our checklist score, going by a public leaderboard and the maker’s pages.
Our checklist
1 source so far
Sources and notes
Public leaderboardMaker’s pagesFree (GPT-6 Luna, s8): 5/10, does the job 1 (arena.ai Creative Writing rank 89 of 412, gpt-6-luna-max, s4), underlying quality 0, value 2 ($0), privacy 1 s9, limits 1. Plus (GPT-6 Sol, the only GPT-6 model in Plus chat, s15): 4/10, does the job 1 (rank 63 of 412, gpt-6-sol-max, s4), underlying quality 0, value 1 ($20/month, s7), privacy 1, limits 1. Score from Free, the higher total. GPT-6.1 Sol, offered to Plus only in Work and Codex, is rank 26 s4 s15; it would also score 5. Top plan: Pro ($100/month) adds GPT-6 Astra in chat, rank 42 s4 s15, which would score 4 (value 0 at that price).
Alternatives
Better or cheaper for agents, from our scores, then the ones we'd also look at. The same kind of product comes first.
- ClaudeApp70$20 a monthScores 80 for coding and hard questions on its free plan, and 70 for writing on Pro. Side by side
- Gemini appApp50$19.99 a monthBest non-Anthropic pick for writing: 70 on AI Pro with Gemini 3.8 Flash, for about the same price. Side by side
- LM Studio (with Bionic agent)App80CodingFreeBest non-Anthropic pick we review for coding: 80 at $0, running open models on your own computer. Side by side
- Claude Opus 5.5BetterText model#1$20 a monthRanked #1 on Arena's Agent board in agents, against 70 here; $20 a month. Side by side
- Gemini 4 ArgonCheaperText model#6$4.99 a monthRanked #6 on Arena's Agent board in agents for $4.99 a month, against $20 a month here. Side by side
Value for money
Where it sits in agents: higher is a better score, further left is cheaper. Green dots are the Best buys.
Evidence
Everything behind the call and the scores, folded so the page stays short. Spotted a wrong price, score or link? Report a mistake.
How we scored this
The agents, coding, research and writing scores come from our own checklist of the product.
Public leaderboardMaker’s pages
Reviews are drafted with help from AI models, including Anthropic's Claude. Model scores come only from outside leaderboards and independent evaluations, by a written rule.
No one paid for this review or gave us free access, and we earn nothing if you sign up. How we score
Sources 15
- s1arena.ai Code Arena: WebDev leaderboard
- s2arena.ai Agent leaderboard
- s3arena.ai Text: Expert leaderboard
- s4arena.ai Text: Creative Writing leaderboard
- s5arena.ai Text: Coding leaderboard
- s6ChatGPT pricing and plan comparison
- s7OpenAI Help: What is ChatGPT Plus?
- s8OpenAI Help: ChatGPT release notes
- s9OpenAI Help: How your data is used to improve model performance
- s10OpenAI Help: Data controls FAQ
- s11Tom's Guide: ChatGPT-6 vs Claude Opus 5.5 with 5 everyday prompts (24 Sep 2026)
- s12OpenAI Help: Using Codex with your ChatGPT plan
- s13OpenAI Help: About ChatGPT Pro tiers
- s14arena.ai Agent Alignment Index
- s15OpenAI Help: GPT-6 and other models in ChatGPT
Something here out of date or wrong? Report a mistake and we will check it against the source.
History 5 changes
Every change to a rating, the price or our call, with the reason. Scores are against what exists on the day, so they drift down as better things ship. Older entries come from before we moved to scores.
- OpenAI's ChatGPT release notes (16 Oct 2026): for Pro users, rolling out, GPT-6 Pro in chat responds faster, uses tools better and can run several agents in parallel on parts of a task. No rating change: ratings use the best model each plan lets you choose, which is unchanged.
- New: Agents ratedNew: Coding ratedNew: Research ratedNew: Writing ratedReviewed: agents, coding, research, writing.
- Worse: Coding downAudit under our rule that each plan is rated on the best model it includes: OpenAI names GPT-6 Sol for Plus chat and GPT-6 Luna for Free, so both checklist lines now use those models instead of the best model in Plus. Coding fell because GPT-6 Sol ranks lower after the Text Coding check, so Free now scores higher. Research and writing keep their ratings but now come from Free. Dropped the Meta Muse test and Instruction Following board, and stopped counting the 24 Sep Tom's Guide test, which did not name the ChatGPT-6 version.
- Better: Coding upRule change: each plan is scored on the best model it lets you choose. OpenAI's help page says Plus includes GPT-6 Astra in ChatGPT Work and Codex, while Plus chat offers only GPT-6 Sol. Agents and coding on Plus are now rated on Astra, so coding returns to 7/10. Research and writing stay on chat models, so they are unchanged. Arena text URLs moved to the /leaderboard/chat/text/ form.
- Like-with-like check: ratings compare Free with Plus only. Confirmed on OpenAI's pricing page that ChatGPT Work is part of Plus. Added top-plan notes for research and writing: Pro adds GPT-6 Astra in chat but would not score higher. No rating changes.
Use it well 4 guides on this site
Pages on this site that help with the jobs above.